Eyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
The DOI matches its Crossref record; 38 of the 42 references it cites were found in a registry or an index.
Abstract
Eyeriss is an accelerator for state-of-the-art deep convolutional neural networks (CNNs). It optimizes for the energy efficiency of the entire system, including the accelerator chip and off-chip DRAM, for various CNN shapes by reconfiguring the architecture. CNNs are widely used in modern AI systems but also bring challenges on throughput and energy efficiency to the underlying hardware. This is because its computation requires a large amount of data, creating significant data movement from on-chip and off-chip that is more energy-consuming than computation. Minimizing data movement energy cost for any CNN shape, therefore, is the key to high throughput and energy efficiency. Eyeriss achieves these goals by using a proposed processing dataflow, called row stationary (RS), on a spatial architecture with 168 processing elements. RS dataflow reconfigures the computation mapping of a given shape, which optimizes energy efficiency by maximally reusing data locally to reduce expensive data movement, such as DRAM accesses. Compression and data gating are also applied to further improve energy efficiency. Eyeriss processes the convolutional layers at 35 frames/s and 0.0029 DRAM access/multiply and accumulation (MAC) for AlexNet at 278 mW (batch size N = 4), and 0.7 frames/s and 0.0035 DRAM access/MAC for VGG-16 at 236 mW (N = 3).
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“such scaling will enable NeuRRAM to deliver higher energy and area efficiency than today’s state-of-the-art edge inference accelerators58–61.”
From A compute-in-memory chip based on resistive random-access memory (2022) · cites this paper as [58]
In the citing paper
“Some studies proposed to implement the argmax() function via hardware128–148, which could be beneficial to reduce the total transistor count and power consumption while at the same time increasing the throughput.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [131]
In the citing paper
“Based on the analysis of the research, CoNNA is able to speed up at least 3.08 times and up to 14.1 times faster than the MIT Eyeriss Accelerator [20].”
From Embedded Intelligence on FPGA: Survey, Applications and Challenges (2021) · cites this paper as [20]
In the citing paper
“massively parallel processing architectures with distributed memory architectures, such as graphic processing units (GPUs) but also increasingly specialized chips for NN inference (e.g., [17]), have shown to be very energy efficient and capable of impressive computational throughput when fully utilized by concurrent algorithms [18].”
From An Introduction to Deep Learning for the Physical Layer (2017) · cites this paper as [17]
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 1×
“Modern CNNs achieve their superior accuracy by building a very deep hierarchy of layers [2]–[5], [7], which transform the input image data into highly abstract representations called feature maps (fmaps).”
BackgroundVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 7×
“We benchmark the chip performance using two publicly available and widely used CNNs: AlexNet [2] and VGG-16 [3].”
MethodInfluentialImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 1×
“[25] demonstrate a CNN processor and only report the theoretical peak throughput along with the power measured on a CNN for the MNIST data set [26], which has storage and computation requirements that are orders of magnitude lower than the state-of-the-art CNNs.”
MethodGoing deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380 · cited 2×
“We have demonstrated an 1000-class image classification task [27] using this system, and a live demo can be found in [29].”
BackgroundMethodRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
Mastering the game of Go with deep neural networks and tree search
Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al. (2016) · Nature · doi:10.1038/nature16961 · Cited by 15,612 · cited 1×
“D EEP learning using convolutional neural networks (CNNs) [1] has achieved unprecedented accuracy on many modern AI applications [2]–[9].”
Background(2010) · International Conference on Machine Learning · Cited by 13,995
Research works printed but not linked: 2
- han, 2015, Learning both weights and connections for efficient neural network, Proc Adv Neural Inf Process Syst, 28, 1135
- chen, 2016, An 1000-Class Image Classification Task Performed by the Eyeriss-Integrated Deep Learning System
Cited by
Items 301–400 of 1,001
Vesti: Energy-Efficient In-Memory Computing Accelerator for Deep Neural Networks
Yin, S., Jiang, Z., Kim, M., Gupta, T., Seok, M., & Seo, J.-S. (2019) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2019.2940649 · Cited by 81
Monolithic 3D Integrated Circuits: Recent Trends and Future Prospects
Dhananjay, K., Shukla, P., Pavlidis, V. F., Coskun, A. K., & Salman, E. (2021) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2021.3051250 · Cited by 100
Lee, E., Han, T., Seo, D., Shin, G., Kim, J., Kim, S., Jeong, S., Rhe, J., Park, J., Ko, J. H., et al. (2021) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2021.3080042 · Cited by 80
AutoDNNchip: An Automated DNN Chip Predictor and Builder for Both FPGAs and ASICs
Xu, P., Zhang, X., Hao, C., Zhao, Y. K., Zhang, Y., Wang, Y., Li, C., Guan, Z., Chen, D., & Lin, Y. (2020) · ACM/SIGDA International Symposium on Field-Programmable Gate Arrays (FPGA) · doi:10.1145/3373087.3375306 · Cited by 101
“Figure 9: The Chip Predictor’s energy prediction error considering the Eyeriss architecture [44]: (a) The energy breakdown for AlexNet’s 1st and 5th convolutional layers and (b) the # of DRAM and SRAM accesses of convolutional layers.”
MethodFeng, C., Gu, J., Zhu, H., Ying, Z., Zhao, Z., Pan, D. Z., & Chen, R. T. (2022) · ACS Photonics · doi:10.1021/acsphotonics.2c01188 · Cited by 98
FPGA-based Deep Learning Inference Accelerators: Where Are We Standing?
Nechi, A., Groth, L., Mulhem, S., Merchant, F., Buchty, R., & Bereković, M. (2023) · ACM Transactions on Reconfigurable Technology and Systems · doi:10.1145/3613963 · Cited by 113
“For ASIC-based DL accelerators, efforts have been made in both industry and academia, where representative ones include TPU [75, 79], ShiDianNao [38], and Eyeriss [26, 27].”
MethodRethinking NoCs for Spatial Neural Network Accelerators
Kwon, H., Samajdar, A., & Krishna, T. (2017) · IEEE/ACM International Symposium on Networks-on-Chip (NOCS) · doi:10.1145/3130218.3130230 · Cited by 74
“The internal dataflow between the PEs is optimized to reuse parameters (input activations, weights, or output activations) that are shared by multiple neurons [3].”
BackgroundMethodInfluentialSelf-citationVIBNN: Hardware Acceleration of Bayesian Neural Networks
Cai, R., Ren, A., Liu, N., Ding, C., Wang, L., Qian, X., Pedram, M., & Wang, Y. (2018) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3173162.3173212 · Cited by 83
“With the success of deep learning, a rising amount of recent works studied the highly parallel computing paradigm and the hardware implementations of neural networks[2, 12, 13, 24, 29, 40, 42, 46].”
BackgroundKim, J.-H., Lee, J., Lee, J., Heo, J., & Kim, J.-Y. (2021) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2020.3039206 · Cited by 78
Google Workloads for Consumer Devices: Mitigating Data Movement Bottlenecks
Boroumand, A., Ghose, S., Kim, Y., Ausavarungnirun, R., Shiu, E., Thakur, R., Kim, D., Kuusela, A., Knies, A., Ranganathan, P., et al. (2018) · ACM SIGPLAN Notices · doi:10.1145/3296957.3173177 · Cited by 176
Moving Deep Learning to the Edge
Véstias, M. P., Duarte, R. P., de Sousa, J. T., & Neto, H. C. (2020) · Algorithms · doi:10.3390/a13050125 · Cited by 73
“Reconfigurable Computing Technology Eyeriss [161] 2018 CNN 46.”
BackgroundGRIP: A Graph Neural Network Accelerator Architecture
Kiningham, K., Levis, P. A., & Ré, C. (2022) · IEEE Transactions on Computers · doi:10.1109/tc.2022.3197083 · Cited by 109
“Our implementation restricts transform to a matrix multiplication, which we implement using a 16 32 weight stationary PE array [23].”
BackgroundMethodReconfigurable Intelligent Surface for Green Edge Inference
Hua, S., Zhou, Y., Yang, K., Shi, Y., & Wang, K. (2021) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2021.3058657 · Cited by 74
HASCO: Towards Agile HArdware and Software CO-design for Tensor Computation
Xiao, Q., Zheng, S., Wu, B., Xu, P., Qian, X., & Liang, Y. E. (2021) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca52012.2021.00086 · Cited by 101
Compact and Computationally Efficient Representation of Deep Neural Networks
Wiedemann, S., Muller, K.-R., & Samek, W. (2020) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2019.2910073 · Cited by 72
“Note that similar energy models have been previously proposed [50], [51].”
Soft Error Resilience of Deep Residual Networks for Object Recognition
Ibrahim, Y., Wang, H., Bai, M., Liu, Z., Wang, J., Yang, Z., & Chen, Z. (2020) · IEEE Access · doi:10.1109/access.2020.2968129 · Cited by 72
S2N2: A FPGA Accelerator for Streaming Spiking Neural Networks
Khodamoradi, A., Denolf, K., & Kastner, R. (2021) · ACM/SIGDA International Symposium on Field-Programmable Gate Arrays (FPGA) · doi:10.1145/3431920.3439283 · Cited by 97
“Frameworks such as FINN [4], fpgaConvNet [48], and Eyeriss [6] provide efficient implementations of this architecture on FPGAs.”
MethodAccelerating Sparse DNN Models without Hardware-Support via Tile-Wise Sparsity
Guo, C., Hsueh, B. Y., Leng, J., Qiu, Y., Guan, Y., Wang, Z., Jia, X., Li, X., Guo, M., & Zhu, Y. (2020) · SC20: International Conference for High Performance Computing, Networking, Storage and Analysis · doi:10.1109/sc41405.2020.00020 · Cited by 69
In-memory computing with emerging nonvolatile memory devices
Cheng, C., Tiw, P. J., Cai, Y., Yan, X., Yang, Y., & Huang, R. (2021) · Science China Information Sciences · doi:10.1007/s11432-021-3327-7 · Cited by 78
FLAT: An Optimized Dataflow for Mitigating Attention Bottlenecks
Kao, S.-C., Subramanian, S., Agrawal, G., Yazdanbakhsh, A., & Krishna, T. (2023) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3575693.3575747 · Cited by 143
“This is because the main operators within attention layers exhibit distinct compute and memory characteristics posing notable bottlenecks on off-chip memory bandwidth compared to CONV and FC.”
BackgroundMethodInfluentialSelf-citationNN-Baton: DNN Workload Orchestration and Chiplet Granularity Exploration for Multichip Accelerators
Tan, Z., Cai, H., Dong, R., & Ma, K. (2021) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca52012.2021.00083 · Cited by 104
“Previous studies for the single-chip accelerator adopted various partition patterns, including stripe [28], rectangle [7], and square [36].”
BackgroundMethodResultBoutros, A., Yazdanshenas, S., & Betz, V. (2018) · ACM Transactions on Reconfigurable Technology and Systems · doi:10.1145/3242898 · Cited by 93
“One might argue that an optimized ASIC design can achieve higher performance by, for example, building custom highly efficient inter-PE network-on-chip such as in Reference [4] or fitting significantly more MACs on-chip [19].”
BackgroundMethodRes-DNN: A Residue Number System-Based DNN Accelerator Unit
Samimi, N., Kamal, M., Afzali‐Kusha, A., & Pedram, M. (2019) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2019.2951083 · Cited by 80
EFLOPS: Algorithm and System Co-Design for a High Performance Distributed Training Platform
Dong, J., Cao, Z., Zhang, T., Ye, J., Wang, S., Feng, F., Li, Z., Liu, X., Song, L., Peng, L., et al. (2020) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca47549.2020.00056 · Cited by 93
Aabrar, K. A., Kirtania, S. G., Liang, F.-X., Gomez, J. T., Jose, M. S., Luo, Y., Ye, H., Dutta, S., Ravikumar, P. G., Ravindran, P. V., et al. (2022) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2022.3142239 · Cited by 74
Wang, L., Liu, S., Zhang, Y., Ding, D., & Yi, X. (2022) · International Journal of Systems Science · doi:10.1080/00207721.2022.2049919 · Cited by 70
“…for the design and analysis of state estimators/filters, i.e. Kalman filtering, H ∞ filtering, l 2 - l ∞ filtering and set-membership filtering (Chen et al., 2017; Chen & Zheng,2015 ; Kan et al., 2018; Ma et al., 2019; Shen, Wang, Wang, & Li, 2020; Shen, Wang, Shen, et al., 2020; Zhao et al.,…”
BackgroundMethodSATA: Sparsity-Aware Training Accelerator for Spiking Neural Networks
Yin, R., Moitra, A., Bhattacharjee, A., Kim, Y., & Panda, P. (2022) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2022.3213211 · Cited by 95
“In general, we follow the gating method used in [29] to omit the computation of MAC and memory read of the filter scratch pad inside the PE when the input is zero.”
MethodLazy Batching: An SLA-aware Batching System for Cloud Machine Learning Inference
Choi, Y., Kim, Y., & Rhu, M. (2021) · 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca51647.2021.00049 · Cited by 69
“While there has been lots of interest in designing energyefficient NPU architectures in isolation [1], [2], [10], [11], [14], [20], [28], [34], [35], [37], [38], [40], [44]-[49], [56], [60], [63]-[65], [72], [80], little attention has been paid in how the M L inference server collects the batched inputs to feed it int…”
BackgroundAngel-Eye: A Complete Design Flow for Mapping CNN onto Customized Hardware
Guo, K., Sui, L., Qiu, J., Yao, S., Han, S., Wang, Y., & Yang, H. (2016) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi.2016.129 · Cited by 79
“Then we give an overview of existing CNN accelerators.”
In-memory computing to break the memory wall*
Huang, X., Liu, C., Jiang, Y., & Zhou, P. (2020) · Chinese Physics B · doi:10.1088/1674-1056/ab90e7 · Cited by 79
A highly stable reliable SRAM cell design for low power applications
Pal, S., Bose, S., Ki, W., & Islam, A. (2019) · Microelectronics Reliability · doi:10.1016/j.microrel.2019.113503 · Cited by 67
An Efficient and Flexible Accelerator Design for Sparse Convolutional Neural Networks
Xie, X., Lin, J. T., Wang, Z., & Wei, J. (2021) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2021.3074300 · Cited by 81
A Survey on the Optimization of Neural Network Accelerators for Micro-AI On-Device Inference
Mazumder, A. N., Meng, J., Rashid, H.-A., Kallakuri, U., Zhang, X., Seo, J.-S., & Mohsenin, T. (2021) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2021.3129415 · Cited by 88
Direct Feedback Alignment With Sparse Connections for Local Learning
Crafton, B., Parihar, A., Gebhardt, E., & Raychowdhury, A. (2019) · Frontiers in Neuroscience · doi:10.3389/fnins.2019.00525 · Cited by 96
“In particular, recent advances in spatial array processors with near-memory computing can achieve significant advantages in both performance and energy-efficiency [26], [27].”
BackgroundNoC-based DNN accelerator: a future design paradigm
Chen, K.-C. J., Ebrahimi, M., Wang, T., & Yang, Y.-C. (2019) · IEEE/ACM International Symposium on Networks-on-Chip (NOCS) · doi:10.1145/3313231.3352376 · Cited by 87
“Thereby, UNPU offers better computational flexibility in calculating various DNN operations, in comparison with other DNN designs such as Eyeriss [7].”
BackgroundMethodInfluentialSayal, A., Nibhanupudi, S. S. T., Fathima, S., & Kulkarni, J. P. (2019) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2019.2939888 · Cited by 72
Li, J., Un, K.-F., Yu, W.-H., Mak, P., & Martins, R. P. (2021) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2021.3095283 · Cited by 80
DSIP: A Scalable Inference Accelerator for Convolutional Neural Networks
Jo, J., Cha, S., Rho, D., & Park, I. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2017.2764045 · Cited by 66
“4 MB [8], but those data are stored into the input and KSP memories in order to reuse them.”
BackgroundMethodInfluentialMATIC: Learning around errors for efficient low-voltage neural network accelerators
Kim, S., Howe, P., Moreau, T., Alaghi, A., Ceze, L., & Sathe, V. (2018) · Design, Automation, and Test in Europe · doi:10.23919/date.2018.8341970 · Cited by 68
“There has been a vast body of work on DNN hardware accelerators [1]-[4], [6], [13], [18] (see Section I), as well as work addressing DNN fault tolerance.”
BackgroundMethodDiffy: a Déjà vu-Free Differential Deep Neural Network Accelerator
Mahmoud, M., Siu, K., & Moshovos, A. (2018) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro.2018.00020 · Cited by 83
Conductive-bridging random-access memories for emerging neuromorphic computing
Cha, J., Yang, S. Y., Oh, J., Choi, S., Park, S., Jang, B. C., Ahn, W., & Choi, S. (2020) · Nanoscale · doi:10.1039/d0nr01671c · Cited by 67
Kadetotad, D. V., Yin, S., Berisha, V., Chakrabarti, C., & Seo, J.-S. (2020) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2020.2992900 · Cited by 68
An In-Memory VLSI Architecture for Convolutional Neural Networks
Kang, M., Lim, S., Gonugondla, S. K., & Shanbhag, N. R. (2018) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2018.2829522 · Cited by 72
“ple, Eyeriss [3], [15], which is a VLSI implementation of AlexNet [1] in a 65 nm CMOS process, reduces the DRAM”
BackgroundMethodInfluentialDeepStore: In-Storage Acceleration for Intelligent Queries
Mailthody, V. S., Qureshi, Z., Liang, W., Feng, Z., de Gonzalo, S. G., Li, Y., Franke, H., Xiong, J., Huang, J., & Hwu, W. (2019) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1145/3352460.3358320 · Cited by 96
“A linear energy model is used to convert these metrics to the energy consumption of each accelerator for the neural network layers of the applications, which is similar to the ones used in [29, 52].”
BackgroundMethodInfluentialHan, D., Im, D., Park, G., Kim, Y., Song, S., Lee, J., & Yoo, H. (2021) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2021.3066400 · Cited by 71
Mapping of local and global synapses on spiking neuromorphic hardware
Das, A., Wu, Y., Huynh, K., Dell'Anna, F., Catthoor, F., & Schaafsma, S. (2018) · Design, Automation, and Test in Europe · doi:10.23919/date.2018.8342201 · Cited by 64
“Although some efforts are made in this direction, the published approaches still remain mostly ad-hoc [5], specific to a given hardware [6] and are often limited to trivial applications [7].”
BackgroundAccPar: Tensor Partitioning for Heterogeneous Deep Learning Accelerators
Song, L., Chen, F., Zhuo, Y., Qian, X., Li, H. H., & Chen, Y. (2020) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca47549.2020.00036 · Cited by 91
“Due to the high demand of computing power in DL applications, we have recently witnessed a phenomenal trend in which the landscape of computing has shifted from general-purpose processors to domain-specific architectures [16–110].”
BackgroundInfluentialSelf-citationDesigning adaptive neural networks for energy-constrained image classification
Stamoulis, D., Chin, T.-W. (., Prakash, A. K., Fang, H., Sajja, S., Bognar, M., & Marculescu, D. (2018) · Proceedings of the International Conference on Computer-Aided Design · doi:10.1145/3240765.3240796 · Cited by 63
“, hardware accelerators [4]).”
Chen, Z., Zhu, H., Ren, E., Liu, Z., Jia, K., Luo, L. Z., Zhang, X., Wei, Q., Qiao, F., Liu, X.-J., et al. (2019) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2019.2937227 · Cited by 63
Safa, A., Corradi, F., Keuninckx, L., Ocket, I., Bourdoux, A., Catthoor, F., & Gielen, G. E. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3109958 · Cited by 75
NAAS: Neural Accelerator Architecture Search
Lin, Y., Yang, M., & Han, S. (2021) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18074.2021.9586250 · Cited by 89
Eyeriss v2: A Flexible and High-Performance Accelerator for Emerging Deep Neural NetworksPreprint
Chen, Y., Emer, J., & Sze, V. (2018) · arXiv (Cornell University) · Cited by 63
Self-citationJiang, L., Kim, M., Wen, W., & Wang, D. (2017) · International Symposium on Low Power Electronics and Design (ISLPED) · doi:10.1109/islped.2017.8009163 · Cited by 80
“Intensive offchip memory accesses seriously limit computing throughput of FPGA [32], [22] and ASIC [7], [1], [2] accelerators, and further shorten the battery lifetime of mobile devices.”
BackgroundInfluentialA novel zero weight/activation-aware hardware architecture of convolutional neural network
Kim, D., Ahn, J., & Yoo, S. (2017) · Design, Automation, and Test in Europe · doi:10.23919/date.2017.7927222 · Cited by 61
“Eyeriss [13] is a CNN accelerator exploiting zero activations to save power.”
BackgroundMethodInfluentialDeep compressive autoencoder for action potential compression in large-scale neural recording
Wu, T., Zhao, W., Keefer, E., & Yang, Z. (2018) · Journal of Neural Engineering · doi:10.1088/1741-2552/aae18d · Cited by 66
“[48] Yu-Hsin Chen, Tushar Krishna, Joel S Emer, and Vivienne Sze.”
BackgroundMethodInfluentialThe Hardware and Algorithm Co-Design for Energy-Efficient DNN Processor on Edge/Mobile Devices
Lee, J., Kang, S., Lee, J., Shin, D., Han, D., & Yoo, H. (2020) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2020.3021397 · Cited by 68
BitBlade: Energy-Efficient Variable Bit-Precision Hardware Accelerator for Quantized Neural Networks
Ryu, S., Kim, H., Yi, W., Kim, E., Kim, Y., Kim, T., & Kim, J. (2022) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2022.3141050 · Cited by 77
Khan, F. H., Pasha, M. A., & Masud, S. (2021) · Micromachines · doi:10.3390/mi12060665 · Cited by 97
“Eyeriss [144] is one of the accelerators based on RS architecture.”
BackgroundDigital Versus Analog Artificial Intelligence Accelerators: Advances, trends, and emerging designs
Seo, J.-S., Saikia, J., Meng, J., He, W., Suh, H.-S., Anupreetham, A., Liao, Y., Hasssan, A., & Yeo, I. (2022) · IEEE Solid-State Circuits Magazine · doi:10.1109/mssc.2022.3182935 · Cited by 77
ACE-CNN: Approximate Carry Disregard Multipliers for Energy-Efficient CNN-Based Image Classification
Shakibhamedan, S., Amirafshar, N., Baroughi, A. S., Shahhoseini, H. S., & TaheriNejad, N. (2024) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2024.3369230 · Cited by 65
“Extensive studies conducted by [40], [42], [43], and [44] have shed light on the computational dynamics of these CNNs. Notably, their findings reveal that a significant portion, approximately 95-99%, of the MACs operations are executed within the convolutional layers.”
BackgroundA 19.4-nJ/Decision, 364-K Decisions/s, In-Memory Random Forest Multi-Class Inference Accelerator
Kang, M., Gonugondla, S. K., Lim, S., & Shanbhag, N. R. (2018) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2018.2822703 · Cited by 70
PIXEL: Photonic Neural Network Accelerator
Shiflett, K., Wright, D., Karanth, A., & Louri, A. (2020) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca47549.2020.00046 · Cited by 77
“A large body of work has recently been introduced focusing on increasing ANN computing speed and power efficiency by developing electronic architectures (such as ASIC and FPGA chips) specifically tailored to improve the data computation or storage ability of the accelerator [8], [9], [7], [10], [11], [12], [13], [14].”
BackgroundHybrid Fusion Paradigm in Advanced Process Monitoring: A Panoramic Review and Future Perspectives
Ali, H., Safdar, R., Liu, J., Manan, T. S. B. A., Hu, G., Rasool, M. H., Yao, Y., & Gao, F. (2025) · Industrial & Engineering Chemistry Research · doi:10.1021/acs.iecr.5c02759 · Cited by 67
MAGMA: An Optimization Framework for Mapping Multiple DNNs on Multiple Accelerator Cores
Kao, S.-C., & Krishna, T. (2022) · 2022 IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca53966.2022.00065 · Cited by 58
“The SG is double-buffered [12], [50] to try and hide the data-fetching latency of the current tile from DRAM/HBM behind the compute latency.”
BackgroundSelf-citationTolerating Soft Errors in Deep Learning Accelerators with Reliable On-Chip Memory Designs
Azizimazreah, A., Gu, Y., Gu, X., & Chen, L. (2018) · 2018 IEEE International Conference on Networking, Architecture and Storage (NAS) · doi:10.1109/nas.2018.8515692 · Cited by 56
±CIM SRAM for Signed In-Memory Broad-Purpose Computing From DSP to Neural Processing
Jain, S., Lin, L., & Alioto, M. (2021) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2021.3092759 · Cited by 60
Designing Hardware for Machine Learning: The Important Role Played by Circuit Designers
Sze, V. (2017) · IEEE Solid-State Circuits Magazine · doi:10.1109/mssc.2017.2745798 · Cited by 64
“Specialized hardware in [42] and [53]–[55] exploits sparse weights for The key metrics for embedded machine learning are accuracy, energy consumption, throughput, and cost.”
BackgroundSelf-citationSim, J., Lee, S., & Kim, L. (2019) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2019.2935251 · Cited by 62
Zhang, H., Chen, D., & Ko, S. (2019) · IEEE Transactions on Computers · doi:10.1109/tc.2019.2936192 · Cited by 65
“A typical deep learning processor architecture [29], [30], [31] contains host interface, DDR memory, on-chip buffer, sequencer, and processing elements (PE), as shown in”
MethodNEUTRAMS: neural network transformation and co-design under neuromorphic hardware constraints
Ji, Y., Zhang, Y., Li, S., Chi, P., Jiang, C., Qu, P., Xie, Y., & Chen, W. (2016) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.5555/3195638.3195663 · Cited by 57
Beyond the Memory Wall: A Case for Memory-Centric HPC System for Deep Learning
Kwon, Y., & Rhu, M. (2018) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro.2018.00021 · Cited by 81
“We believe our methodology provides accurate estimations without losing fidelity due to the following two reasons: (1) as DNN computation and memory accesses have high data locality with highly deterministic dataflow, existing designs [42], [46], [47] primarily employ a lightweight FSM or microcontrollers to orchestra…”
BackgroundMethodComprehensive Evaluation of Supply Voltage Underscaling in FPGA on-Chip Memories
Salami, B., Ünsal, O., & Kestelman, A. C. (2018) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro.2018.00064 · Cited by 66
Breaking High-Resolution CNN Bandwidth Barriers With Enhanced Depth-First Execution
Goetschalckx, K., & Verhelst, M. (2019) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2019.2905361 · Cited by 69
Benchmarking Contemporary Deep Learning Hardware and Frameworks: A Survey of Qualitative Metrics
Dai, W., & Berleant, D. (2019) · 2019 IEEE First International Conference on Cognitive Machine Intelligence (CogMI) · doi:10.1109/cogmi48466.2019.00029 · Cited by 43
“Furthermore, the Eyeriss achieved 125.9 images/joule (with batch size N=4) [19].”
InfluentialDNNExplorer: a framework for modeling and exploring a novel paradigm of FPGA-based DNN accelerator
Zhang, X., Ye, H., Wang, J., Lin, Y., Xiong, J., Hwu, W., & Chen, D. (2020) · doi:10.1145/3400302.3415609 · Cited by 124
Machine-learning-assisted and real-time-feedback-controlled growth of InAs/GaAs quantum dots
Shen, C., Zhan, W., Xin, K., Li, M., Sun, Z., Cong, H., Xu, C., Tang, J., Wu, Z., Xu, B., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-47087-w · Cited by 64
Energy-Efficient Convolution Architecture Based on Rescheduled Dataflow
Jo, J., Kim, S., & Park, I. (2018) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2018.2840092 · Cited by 57
“Since the previous accelerators, [7] and [8], access on-chip memories frequently, enormous energy consumption is induced by the on-chip memory access.”
BackgroundMethodInfluentialGilan, A. A., Emad, M., & Alizadeh, B. (2019) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2019.2922372 · Cited by 61
“In [2], a chip has been fabricated that can efficiently accelerate CNNs with an extremely low power consumption.”
VIBNN: Hardware Acceleration of Bayesian Neural Networks
Cai, R., Ren, A., Liu, N., Ding, C., Wang, L., Qian, X., Pedram, M., & Wang, Y. (2018) · ACM SIGPLAN Notices · doi:10.1145/3296957.3173212 · Cited by 78
IMCE: Energy-efficient bit-wise in-memory convolution engine for deep neural network
Angizi, S., He, Z., Parveen, F., & Fan, D. (2018) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/aspdac.2018.8297291 · Cited by 75
“It applies kernels (filters) by performing high-dimensional convolutions and then produces the features (Output fmaps) for the next layer [13].”
BackgroundDVAFS: Trading computational accuracy for energy through dynamic-voltage-accuracy-frequency-scaling
Moons, B., Uytterhoeven, R., Dehaene, W., & Verhelst, M. (2017) · Design, Automation, and Test in Europe · doi:10.23919/date.2017.7927038 · Cited by 60
“First, hardware implementations of neural networks can be made more efficient by exploiting algorithm-level parallelism, as well as temporal and spatial data-locality [15]–[19].”
MethodRao, C., Yu, H., Wan, H., Zhou, J., Zheng, Y., Wu, M., Ma, Y., Chen, A., Yuan, B., Zhou, P., et al. (2022) · ACM Transactions on Graphics · doi:10.1145/3550454.3555505 · Cited by 61
Shafique, M., Hafiz, R., Javed, M. U., Abbas, S., Sekanina, L., Vasicek, Z., & Mrazek, V. (2017) · 2017 IEEE Computer Society Annual Symposium on VLSI (ISVLSI) · doi:10.1109/isvlsi.2017.124 · Cited by 53
“Therefore, the design of highly energy-efficient hardware accelerators for compute-intensive CNNs (or their kernels) has received a lot of interest [5].”
BackgroundMethodInfluentialE2-Train: Training State-of-the-art CNNs with Over 80% Energy SavingsPreprint
Wang, Y., Jiang, Z., Chen, X., Xu, P., Zhao, Y. K., Lin, Y., & Wang, Z. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.13349 · Cited by 69
“Also, L ( W, G ) denotes the prediction loss, and C ( W, G ) is calculated by accumulating the computational cost (FLOPs) of the layers that are selected.”
BackgroundMethodInfluentialConfigurable Multi-directional Systolic Array Architecture for Convolutional Neural Networks
Xu, R., Ma, S., Wang, Y., Chen, X., & Guo, Y. (2021) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3460776 · Cited by 56
“Flexflow [27] and Eyeriss [2] propose new dataflows and change the systolic architecture.”
BackgroundPointAcc: Efficient Point Cloud Accelerator
Lin, Y., Zhang, Z., Tang, H., Wang, H., & Han, S. (2021) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 54
Distilling Bit-level Sparsity Parallelism for General Purpose Deep Learning Acceleration
Lü, H., Chang, L., Li, C., Zhu, Z., Lu, S., Liu, Y., & Zhang, M. (2021) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1145/3466752.3480123 · Cited by 64
“Exploited Support bit parallel Eyeriss[10], N/A 16b No DaDianNao[12]Cambricon A-/W-value 16b No -”
BackgroundInterpretability of deep neural networks: A review of methods, classification and hardware
Antamis, T., Drosou, A., Vafeiadis, T., Nizamis, A., Ioannidis, D., & Tzovaras, D. K. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128204 · Cited by 70
ReRAM-Based In-Memory Computing for Search Engine and Neural Network Applications
Halawani, Y., Mohammad, B. S., Lebdeh, M. A., Al‐Qutayri, M., & Al-Sarawi, S. F. (2019) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2019.2909317 · Cited by 59
Imperceptible Misclassification Attack on Deep Learning Accelerator by Glitch Injection
Liu, W., Chang, C.-H., Zhang, F., & Lou, X. (2020) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18072.2020.9218577 · Cited by 57
Drawing Early-Bird Tickets: Towards More Efficient Training of Deep NetworksPreprint
You, H., Li, C., Xu, P., Fu, Y., Wang, Y., Chen, X., Baraniuk, R. G., Wang, Z., & Lin, Y. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.11957 · Cited by 128
“We measure the training energy costs from realdevice operations as the energy cost consists of both computational and data movement costs, the latter of which is often dominant but can not captured by the commonly used metrics, such as the number of FLOPs (Chen et al., 2017), we evaluate the proposed techniques agains…”
AccelTran: A Sparsity-Aware Accelerator for Dynamic Inference With Transformers
Tuli, S., & Jha, N. K. (2023) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2023.3273992 · Cited by 74
Integrating silicon photonics with complementary metal–oxide–semiconductor technologies
Wan, Y., He, W., Jaussi, J. E., Liao, L., Pan, D. Z., Bowers, J. E., & Rong, H. (2025) · Nature Reviews Electrical Engineering · doi:10.1038/s44287-025-00223-0 · Cited by 77
“Traditional digital electronics, such as GPUs and ASICs 141,142 , are facing mounting challenges, such as millisecond-level latency, high energy consumption, excessive heat production and elevated interconnect costs 143,144 .”
Shah, N., Chaudhari, P., & Varghese, K. (2018) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2815085 · Cited by 58
Ueyoshi, K., Ando, K., Hirose, K., Takamaeda-Yamazaki, S., Hamada, M., Kuroda, T., & Motomura, M. (2018) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2018.2871623 · Cited by 57
“) DNN accelerators [17], [23]–[25], using LUT-based or linear quantization, with this paper.”
Fast and Efficient Convolutional Accelerator for Edge Computing
Ardakani, A., Condo, C., & Gross, W. J. (2019) · IEEE Transactions on Computers · doi:10.1109/tc.2019.2941875 · Cited by 53
“52 mm(2) (1,852 kgates) and tested on AlexNet and VGGNet-16 [8], [24].”
BackgroundMethodEverything Leaves Footprints: Hardware Accelerated Intermittent Deep Inference
Kang, C.-K., Mendis, H. R., Lin, C., Chen⋆, M.-S., & Hsiu, P.-C. (2020) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2020.3012217 · Cited by 53
“However, even the most power-optimized DNN hardware accelerators can still consume tens to hundreds of mW of power [18], [23], which is several orders of magnitude higher than that typical ambient energy harvesters can provide [25], [34].”
BackgroundExploring the Accuracy – Energy Trade-off in Machine Learning
Brownlee, A. E. I., Adair, J., Haraldsson, S. O., & Jabbo, J. (2021) · IEEE/ACM International Workshop on Genetic Improvement (GI) · doi:10.1109/gi52543.2021.00011 · Cited by 80
Liu, Z., Ren, E., Qiao, F., Wei, Q., Liu, X.-J., Luo, L., Zhao, H., & Yang, H. (2020) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2020.2984161 · Cited by 57
Evaluation Method of Deep Learning-Based Embedded Systems for Traffic Sign Detection
Lopez-Montiel, M. A., Orozco-Rosas, U., Sánchez-Adame, M., Picos, K., & Montiel, O. (2021) · IEEE Access · doi:10.1109/access.2021.3097969 · Cited by 61
“[57] generated hardware accelerators both in the hardware and software fields to improve the performance and take advantage of the use of DNNs and CNNs and optimizing and maximizing data reuse, data movement, and statistical exploitation.”
Metadata
- Authors
- Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V.
- Affiliations
- Massachusetts Institute of Technology (US)
- Georgia Institute of Technology (US)
- Nvidia (United States) (US)
- Venue
- IEEE Journal of Solid-State Circuits
- Published
- 2017-01
- Volume
- 52
- Issue
- 1
- Pages
- 127-138
- ISSN
- 0018-9200 (ISSN-L)
- 1558-173X
- OpenAlex
- W2289252105
- Semantic Scholar
- ffdaa12ef011de9dbf43be46d45a3abcc8288965
- Cited by
- 3,247
- References
- 42
- Citations by year
- 2026: 218 · 2025: 313 · 2024: 294 · 2023: 373 · 2022: 365 · 2021: 450 · 2020: 427 · 2019: 354 · 2018: 302 · 2017: 118
- FWCI
- 182.12
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- IEEE Journal of Solid-State Circuits (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Keywords
- reconfigurable accelerators, deep convolutional neural networks, energy efficiency, spatial architecture, data movement reduction, data compression, processing elements, AlexNet, VGG16, convolutional layers
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.